Abstract
The pursuit of automating contracts has persisted for several decades, driven by efforts to reduce human intervention across the entire contract life cycle. While the notion of minimizing human oversight in contract management is not a recent development, the emergence of artificial intelligence has revolutionized this process in such a profound manner that might affect contract law.AI is pushing the boundaries of contract formation, representing one of the emerging domains where its impact is becoming increasingly evident. However, the enforceability of these AI- negotiated contracts remains one of the least explored areas. The challenges posed to contract law by the dehumanization of the contract formation phase, replaced by the interaction of algorithms or the 'meeting of algorithms', are recognized but only barely addressed by scholars.
This research aims to analyze this phenomenon from the perspective of the so-called algorithmic contracts. For this purpose, and narrowing down even more AI-negotiated contracts, I define an algorithmic contract as a contractual agreement where both involved parties engage in decision-making processes operating an algorithmic system rooted in artificial intelligence.
The existing literature on this topic reveals divergent perspectives. Some view these AI-based systems as mere tools, attributing to them limited capabilities in overall reasoning and decision- making. In contrast, some argue for recognizing these AIs as legal agents. However, the fact that AI systems cannot be categorized as legal agents leads others to propose the recognition of legal personality to them, an option that is widely perceived as excessive. While the theory of mere tools appears fitting, it would be inaccurate to assume, as implied by the oversimplified version of this theory, that AI systems are incapable of taking on decision-making functions. What sets algorithmic contracts apart is the distinctive and novel aspect where computer programs, during the formation of such contracts, assume decision-making functions that were traditionally carried out by humans. To date, none of these approaches has provided a solution to the so-called 'contracting problem'.
To construct a comprehensive theory of AI-driven contract formation, it is essential to reassess the foundational elements of contract law. This research seeks to establish, from the perspective of English contract law, a robust normative legal framework that effectively accommodates and clarifies the algorithmic contract formation phase. From a 'coherentist' approach, my hypothesis postulates that by dissecting this phenomenon into its fundamental phases or components, each of these parts can be integrated into the framework of contract law, thereby facilitating the integration of algorithmic contracts in their entirety. In pursuit of this objective, it is paramount to scrutinize its communicative role to delineate the moment of algorithmic contract formation. Accordingly, the research reconsiders how this process can occur without the explicit awareness of the system operator. For this purpose, it is essential to find a mechanism to attribute to the contracting parties both knowledge of the conclusion of the contract and of its content at the time of contract formation. Ultimately, the challenge lies in establishing the legal basis for equating algorithmic interactions with the "meeting of the minds".
| Date of Award | 1 Aug 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Mateja Durovic (Supervisor) & Yael Lifshitz (Supervisor) |